At 9:00 a.m. Eastern today, April 7, 2026, the EIA published its Short-Term Energy Outlook and inserted a piece of data that will not make the financial press but will reshape grid planning through 2027: U.S. electricity demand from both residential and commercial sectors is expected to grow 3% relative to last summer, with commercial demand leading at 6% growth this summer and accelerating to 6% again in summer 2027. That is not a forecast artifact or seasonal noise. That is a load statement. And it lands precisely as the Strait of Hormuz is closed enough that Iraq, Saudi Arabia, Kuwait, UAE, Qatar, and Bahrain have collectively shut in 7.5 million barrels per day of crude production in March, rising to 9.1 million b/d in April. The message is unmistakable: the grid cannot burn its way to stability anymore.
The Hormuz blockade has done what structural energy analysis could only theorize about before: made the gas supply ceiling visible and unmovable. U.S. LNG export facilities are running at near-peak capacity, exporting almost 18 billion cubic feet per day in March, close to the record set in December 2025. That capacity is a hard ceiling. The Henry Hub-to-Europe spread has widened because LNG cannot flow any faster through U.S. terminals. Brent crude has climbed to $103 per barrel on average in March and is expected to peak at $115 per barrel in Q2 2026 before easing as OPEC shut-ins gradually abate. Higher oil prices make gas-to-power more expensive. Expensive gas makes storage and solar dispatch more valuable. The EIA is not saying that directly. But the April STEO says it in load numbers. The grid has to build its way out.
That build is already underway, and it is massive. Developers plan to add 86 GW of new utility-scale electric generating capacity to the grid in 2026—a record if realized. Solar makes up 51% of planned additions (43.4 GW), battery storage 28% (24 GW), and wind 14%. The battery storage number is the inflection: developers plan to add 24 GW this year, compared with a record 15 GW added in 2025, a 56.7% increase year-over-year. Utility-scale battery storage capacity will close out 2026 at nearly 65 GW, more than double the 17 GW installed in Q1 2024 just two years prior. The single largest solar-plus-storage project expected online in 2026 is Tehuacana Creek 1 in Texas—837 MW of solar PV plus 418 MW of battery storage. That is not a pilot project. That is a grid asset. And it is one of dozens.
What made this moment possible is straightforward: the capital markets finally stopped treating battery storage as a hedge against future uncertainty and started treating it as an answer to present demand. In 2025, the U.S. electric power sector generated 4,275 billion kilowatt-hours of electricity, up 2.9% from 2024; renewable energy provided over a quarter of that generation, up 11% year-over-year. Solar, wind, and batteries added over 55 GW of new capacity in 2025 while the net total from fossil fuels and nuclear was less than 1 GW. That is not a transition in the press release. That is a transition in the flow of capital and material. Meanwhile, EIA's own administrator, Tristan Abbey, made explicit in the January STEO that electricity demand is being driven by "large computing facilities, including data centers"—not by residential electrification or industrial expansion, but by AI load. The data center demand vector was always there; what changed is that it is now large enough and visible enough that it overrides the usual seasonal and demographic drivers. And Hormuz has made it impossible to paper over with gas.
But the benefit is deeply unequally distributed. Three states account for roughly 80% of planned 2026 battery storage capacity: Texas at 53% (12.9 GW), California at 14% (3.4 GW), and Arizona at 13% (3.2 GW). That is not a grid. That is a regional concentration risk that reads like a strategic vulnerability. ERCOT (which is Texas) is forecast to increase electricity generation by 7.3% between 2025 and 2026—an unusually steep ramp. The transmission infrastructure to move that power from those three states to load centers in the Northeast, Midwest, and Southeast does not exist yet. The PJM board approved an $11.8 billion transmission expansion plan on February 12, 2026, but Dominion Energy's Virginia utility landed only $4.8 billion of those projects, and the state-level siting battles have only just begun this spring. The grid is building storage faster than it can build transmission to distribute it. That is not a feature of the plan; it is a fault line.
Our read: The April 7 STEO is a structural signal that the U.S. electricity system has crossed a threshold from gas-dominated marginal economics to storage-constrained capacity dynamics. The Hormuz crisis did not cause this shift—it accelerated visibility of one that was already inevitable given data center load growth. What the STEO actually reveals is that 2026 is now a validation year, not a forecast year. If the 24 GW of battery storage does come online on schedule, the grid absorbs the commercial demand surge with minimal fossil fuel burning and positions itself for the 2027 acceleration. If permitting delays, interconnection queue backlog, or the FERC large-load ruling create friction, summer 2026 becomes a stress test that proves the regional concentration problem is real and unsolvable without transmission. The commercial demand number (6% growth) is the key metric to watch. That is not seasonal. That is not demographic. That is compute load, and it will not reverse. Three signals would prove this read wrong: first, if EIA's Annual Energy Outlook webinar tomorrow (April 8) walks back the commercial demand forecast significantly; second, if FERC's April 30 large-load ruling narrows the 20 MW threshold or allocates interconnection costs in a way that delays projects more than six months; third, if summer 2026 peak demand is met without incremental storage coming online, suggesting the entire demand forecast was overcooked. Any of those would mean the grid can still burn through this transition. We do not think it can.
Watch four things with explicit calendars and thresholds. First: FERC's large-load interconnection final rule hits April 30, 2026—in 23 days. The 20 MW applicability threshold, cost allocation, and hybrid facility treatment will determine whether the 24 GW storage pipeline lands on schedule or backs up by quarters. Second: ERCOT and EIA releases for May–September 2026 summer reserve margins and peak load realization. If the commercial demand forecast holds and storage does not arrive, you will see reserve margins compress below 10% in Q3, which has historically triggered emergency procedures. Third: Dominion Virginia HVDC siting decisions across Virginia, Pennsylvania, West Virginia, and Maryland, with intervention deadlines this spring and state commission decisions expected by fall 2026. If that $4.8 billion in transmission clears, it solves a piece of the Northeast regional crunch. If it does not, the grid has a topology problem that no amount of storage at the source can overcome. Fourth: the first reports of 2026 battery storage installations versus plan, coming in May and June data releases. If actual vs. planned installations diverge by more than 20%, the entire capacity build thesis breaks and Q3 becomes dangerous.